Optimizing neural-network learning rate by using a genetic algorithm with per-epoch mutations.
Yasusi KanadaPublished in: IJCNN (2016)
Keyphrases
- learning rate
- genetic algorithm
- neural network
- adaptive learning rate
- mutation rate
- hidden layer
- activation function
- learning algorithm
- training algorithm
- convergence rate
- rapid convergence
- convergence speed
- artificial neural networks
- error function
- feed forward neural networks
- fitness function
- fuzzy logic
- differential evolution
- evolutionary process
- weight vector
- evolutionary algorithm
- multi objective
- fuzzy neural network
- convergence theorem
- back propagation
- evolutionary computation
- multilayer neural networks
- simulated annealing
- bp neural network algorithm
- feedforward neural networks
- mutation operator
- control parameters
- genetic algorithm ga
- neural nets
- recurrent neural networks
- neural network model
- binary decision tree
- optimization method